Individualized Prediction of Task Performance Decline Using Pre-Task Resting-State Functional Connectivity.

Area of Science:

• Neuroscience
• Cognitive Science
• Psychology

Background:

• Mental fatigue, particularly the time-on-task (TOT) effect, significantly impairs daily activities and performance.
• Individual differences in task efficiency and brain activity make predicting fatigue-related decline at the individual level challenging.
• Current group-level analyses lack generalizability for personalized fatigue prediction.

Purpose of the Study:

• To investigate the feasibility of using pre-task resting-state functional connectivity (FC) to predict individual fatigue-related performance degradation.
• To develop and validate a data-driven framework for individualized prediction of fatigue-related behavioral impairment.

Main Methods:

• Utilized a cross-validated, data-driven analysis framework on resting-state functional connectivity (FC) data from 37 healthy subjects.
• Estimated individual behavioral impairment using metrics like ∆RT and TOTslope over a 15-min sustained attention task across three sessions.
• Employed a connectome-based prediction model using pre-task FC features to predict performance deterioration.

Main Results:

• Confirmed significant populational TOT-related performance declines with substantial inter-individual variability.
• Achieved high accuracy in predicting individual TOT-related behavioral impairment metrics using pre-task neuroimaging features.
• Identified key FC features within/between frontal, temporal, and parietal areas contributing to individualized predictions.

Conclusions:

• Pre-task resting-state FC is a viable predictor of individual fatigue-related performance decline.
• The developed framework offers a promising approach beyond conventional correlation/classification methods.
• This methodology may lead to practical techniques for preventing performance decrements in real-world scenarios.